INDEX

Explanations

meaning, purpose, nature of existence

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

10,000 prompts, 128 tokens each

Dataset (Dashboard)

lmsys/lmsys-chat-1m

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ubit

-0.11

orra

-0.10

lesc

-0.09

AIT

-0.09

ENE

-0.09

 Karlov

-0.09

uhl

-0.09

Ð¾ÑĢÐ¾Ð²

-0.08

á»ĵng

-0.08

Ø§Ø°

-0.08

POSITIVE LOGITS

 meaning

0.28

 purpose

0.24

 Meaning

0.22

meaning

0.20

 Purpose

0.19

 questions

0.19

 nature

0.18

 existence

0.18

 ultimate

0.18

Purpose

0.17

Activations Density 0.091%